Reelsmith

Create short-form vertical video packages, preview reels, narrated reels, and AI-video workflows from ideas, articles, updates, or source material. Use when...

This skill's code and SKILL.md implement what 'Reelsmith' claims to do, but the package metadata omits important runtime requirements. Before installing or running it: 1) Request corrected metadata that explicitly lists required env vars (OPENAI_API_KEY and, if you plan to use the LTX backend, LTX_API_KEY), required binaries (ffmpeg) and Python dependencies (openai, requests). 2) Verify you trust the external endpoint https://api.ltx.video and confirm its privacy/usage terms before supplying an LTX API key. 3) Run the scripts in a controlled environment (sandbox/container) so ffmpeg and subprocess calls can't access unexpected host files. 4) Only provide API keys you intend to use for this capability (consider scoped keys or separate service accounts). If the publisher cannot or will not clarify the missing declarations, treat the package as untrusted until those inconsistencies are resolved.

Type: OpenClaw Skill Name: reelsmith Version: 0.2.0 The reelsmith skill is a legitimate tool for generating short-form video content packages, including scripts, captions, and video previews. It utilizes local ffmpeg processing (e.g., scripts/make_visual_reel_preview.py) and external APIs for text-to-speech (scripts/openai_tts.py) and AI video generation (scripts/ltx_text_to_video.py). The code follows safe practices, such as using subprocess lists instead of shell strings and utilizing ffmpeg's textfile parameter to avoid injection vulnerabilities, and its behavior aligns strictly with the documented purpose.